1. Introduction
Among the registered manuscripts containing the Greek text of the Gospel of John, there are a total of five papyrus fragments and three parchment fragments known as ‘hermeneia’ manuscripts, that is, fragments containing a certain passage from the Gospel of John, below which occurs the word ἑρμηνεία, centred on the page, which is then followed by a kind of enigmatic comment or note on the biblical citation.Footnote 1 All of these manuscripts follow precisely this tripartite pattern: (1) some text of John, (2) the word ἑρμηνεία and (3) a brief comment. To give just one example, I reproduce here the verso of P.Monts. Roca 83Footnote 2 (formerly P.Barc. 83), also known as P80:
[Text of John 3.34]
ἑρμηνίαFootnote 3
ἀληθῆ ἐστιν τά λ̣[ελαλημένα]
παρ' αὐτοῦ ἐάν σ̣[υ ἐν αὐτοῖς]
ὠφεληθήσῃ.
Thus, the comment appears to be a statement that expresses something further about the phrase ‘speaks the words of God’ (τὰ ῥήματα τοῦ θεοῦ λαλεῖ) that occurs in the Johannine citation. In addition to the Greek comments in these eight manuscripts, there are also comments in Coptic alongside the Greek in P.Berlin 11914 (P63) and Paris, BnF Copte 156, a Greco-Coptic manuscript discovered at Antinoe and published by Walter E. Crum in 1904.Footnote 4 There are also ἑρμηνεία comments in Latin in Codex Sangermanensis (GA 0319), as well as in manuscripts in Armenian and Georgian.Footnote 5 All extant ἑρμηνεία manuscripts, of which only a handful are known to us today, are by definition ‘non-continuous’ manuscripts (or fragments thereof), in that they were not originally written out as complete and continuous (i.e. unbroken or uninterrupted) copies of whole books.Footnote 6
The function of these comments is anything but clear. Bruce M. Metzger, Stanley E. Porter and, most recently, Wally V. Cirafesi and Kevin W. Wilkinson have all written significant articles on the ἑρμηνεῖαι.Footnote 7 Drawing on J. Rendel Harris' work on sortes sanctorum and the Greek–Latin ἑρμηνεῖαι in Codex Bezae (GA 05) and Codex Sangermanensis (GA 0319), Metzger argues that these special manuscripts were likely used for the purpose of divination and not as a reading copy of the Gospel.Footnote 8 According to Metzger, the ἑρμηνεῖαι were oracles disconnected from the biblical text above.
Other scholars, such as Porter, disagree with the theory that these are oracular statements. According to Porter, the ἑρμηνεῖαι are ‘biblically motivated and connected reflections on the biblical text’, or at least individual parts thereof.Footnote 9 In a similar line of argument, Cirafesi, highlighting the bilingual character of these manuscripts, suggests that ‘ἑρμηνεῖαι are interpretive comments (loosely understood) that functioned as liturgical tools to facilitate early Christian worship services needing to accommodate the use of two languages within a particular community’.Footnote 10 In support of this thesis, one may also point to the occurrence of the ‘summary notes’ in P.Bodmer viii (P72), where it appears that a Coptic scribe was responsible for drawing attention to certain themes in the margin,Footnote 11 or the Coptic glosses in Old Fayyumic in P.Beatty vii (Isaiah),Footnote 12 not to mention anything of the Greco-Coptic lectionaries and various Greco-Coptic New Testament diglots.Footnote 13 Such phenomena demonstrate clearly that Coptic and Greek co-existed within many Coptic Christian communities and so Cirafesi's theory concerning liturgical contexts and the need to accommodate the use of more than one language is appealing.
In a forthcoming essay, Kevin W. Wilkinson argues that the ἑρμηνεῖαι were ‘an aid to bibliomancy’, and that the comments are clearly related to the gospel passages that they accompany (contra Metzger). According to Wilkinson, ‘[a]nyone wishing to inquire into his or her fate would arrive by some means at a passage of John and then consult the accompanying “interpretation,” which translated the language and/or content of the biblical text into an oracular prediction or command’.Footnote 14 Wilkinson's treatment of the ἑρμηνεῖαι provides much of the clarity necessary for understanding the structure of the oracular system. However, while the questions about the very nature and purpose of the ἑρμηνεῖαι have not been fully answered, such lines of inquiry are outside the scope of this study.
Now that I have given a brief discussion of the ἑρμηνεῖαι I would like to turn to the primary purpose of this paper. In the summer of 2013, while examining various manuscripts in the Beinecke Rare Book and Manuscript Library of Yale University, I came across P.CtYBR inv. 4641, a previously unpublished Coptic parchment codex leaf, and identified it as a copy of the Gospel of John in the Sahidic dialect containing portions of chapter 3. I further realised that this manuscript contains ἑρμηνεῖαι on both the flesh and hair sides, arranged in the same tripartite structure as all other Johannine ἑρμηνεία manuscripts (i.e. citation of John, the word ἑρμηνεία, a brief comment). As such, P.CtYBR inv. 4641 represents the first known example of a Coptic-only manuscript with both the text of John and the ἑρμηνεῖαι in Coptic; the other examples that do contain Coptic comments are bilingual.Footnote 15 Thus, the Yale fragment has much significance for discussions about the ἑρμηνεία manuscripts, their origin, influences and functions. Below, I publish P.CtYBR inv. 4641 by discussing relevant issues pertaining to the manuscript, offering a transcription of its text, and recording variants of special interest.Footnote 16
2. The Manuscript
P.CtYBR inv. 4641Footnote 17 14.6 x 9.1 cm 5th–7th c. ce
Provenance Unknown
Yale University purchased the manuscript in 1996 from Gallery Nefer, Zurich, owned by Frieda Tchacos Nussberger, the famous Zurich antiquities dealer who was instrumental in bringing the Gospel of Judas to light. It was inventoried under the genre of a ‘literary work’ but its contents remained unidentified until now. The fragment measures 14.6 cm high × 9.1 cm wide, and hair and flesh are distinguishable by colour. There are eleven lines of text on the flesh side and fifteen lines on the hair side. The fragment is from the bottom portion of a codex folio; the bottom left margin (flesh) and bottom right margin (hair) are preserved. Original dimensions cannot be reconstructed with any precision, since the text is not written in a continuous fashion. However, we may tentatively suggest that the upper part of the hair side (now lost) did not contain much text, since it is separated from the text of the flesh by only eleven or so words (by reconstruction). If this estimation is correct, then the original size of the codex must have been relatively small, perhaps falling within Turner's Categories 9 or 11.Footnote 18
The text is arranged in a single column and written in a very neat and elegant hand. The script is unimodular (or biblical majuscule), and the letters are strictly bilinear and lack decoration except for a few very light finials or return strokes on the tips of some letters and serifed 
's. The script is upright, with three-stroke 
, wide 
, short 
, and tall ϥ and ρ.Footnote 19 The thickness of strokes is virtually uniform; horizontal strokes are at times only slightly thinner than the vertical strokes, which is less common for Coptic manuscripts with wide . The only form of punctuation is the middle dot, which occurs twice (flesh l. 9, hair l. 12). Surprisingly, supralinear strokes (of both the connective and single-letter types) are completely absent.Footnote 20 The hand of P.CtYBR inv. 4641 may be compared to P.Monts. Roca ii 8 (2 Samuel).Footnote 21 It is also very similar to BM Or. 6696 (Psalter) and BM Or. 6697 (Acts; Horner's ‘14’), although the contrast between thick and thin strokes is more pronounced in these latter manuscripts.Footnote 22
We have no sufficient evidence that would allow us to secure a precise date for this manuscript. Unlike Greek palaeography, which follows a fairly established set of criteria as well as a general understanding of the development of Greek literary hands, Coptic palaeography has been established on criteria that are both circular and unreliable, and many editors of Coptic texts leave the dating open, an approach I myself have taken in the past. That said, however, I suggest that, adopting wide parameters, the manuscript was written before the Arab conquest (ca. 640) on account of its (1) relatively small size, (2) single-column format,Footnote 23 (3) uniform thickness of strokes and (4) lack of decorations and other signs of stylistic development. Indeed, some early Coptic manuscripts do contain ornamentations and some late manuscripts lack them. For the most part, however, early Coptic manuscripts are without heavy decorations, and when all three features above are considered together, it increases the probability that our manuscript is earlier rather than later. Thus, I tentatively propose a date of ca. 5th–7th century ce.
3. The Text
For convenience, restorations of lacunae and word division are based on the edition of Horner.Footnote 24 Punctuation, tremata and supralinear strokes have also been reproduced from Horner. We have compared our transcript with the edition of Horner (= H; and variants), Quecke's edition of P.Palau Rib. Inv.-Nr. 183 (= sa 1), as well as the variants of Chester Beatty Library Cpt. 813 (= sa 4), Chester Beatty Library Cpt. 814 (= sa 5) and Pierpont Morgan M569 (= sa 9), which are cited by H. Quecke.Footnote 25 These are cited below in the apparatus and commentary at relevant points of discussion.
4. Critical Apparatus
Flesh
1
![ϫⲉ]ⲕⲁⲥ̣: ϫⲉⲕⲁⲁⲥ](data:image/gif;base64,R0lGODlhAQABAIAAAMLCwgAAACH5BAAAAAAALAAAAAABAAEAAAICRAEAOw==)
sa 45
omits BnF Copte 129(9) (Horner's 91 and 133)7
sa 9
![ϫⲉ]ⲕⲁⲥ̣: ϫⲉⲕⲁⲁⲥ](https://static.cambridge.org/binary/version/id/urn:cambridge.org:id:binary:46467:20160415020217788-0178:S0028688513000374_inline33.gif?pub-status=live){kind=small}
sa 4![ⲡⲉⲧ]ⲉⲛ̄ϥⲡⲓⲥⲧⲉ[ⲩⲉ: ⲡⲉⲧⲉ-](https://static.cambridge.org/binary/version/id/urn:cambridge.org:id:binary:52362:20160415020217788-0178:S0028688513000374_inline34.gif?pub-status=live){kind=small}
omits BnF Copte 129(9) (Horner's 91 and 133)
sa 9Hair
1
sa 12
sa 43
: omits sa 95–6
: omits sa 16
H // sa 98
: omits H9
sa 1
![ⲛ]ⲟⲩⲟ: ⲉⲟⲩⲉ](https://static.cambridge.org/binary/version/id/urn:cambridge.org:id:binary:58847:20160415020217788-0178:S0028688513000374_inline36.gif?pub-status=live){kind=small}
sa 1![ⲛⲉⲩⲃ]ⲏⲩⲉ: ⲛⲉⲩⲃⲏⲟⲩⲉ](https://static.cambridge.org/binary/version/id/urn:cambridge.org:id:binary:76213:20160415020217788-0178:S0028688513000374_inline37.gif?pub-status=live){kind=small}
sa 4
: omits sa 9![ⲁ]ⲩⲱ ⲙⲉϥⲉⲓ ϣ[ⲁ | ⲡⲟⲩⲟⲉⲓⲛ]](https://static.cambridge.org/binary/version/id/urn:cambridge.org:id:binary:46178:20160415020217788-0178:S0028688513000374_inline39.gif?pub-status=live){kind=small}
: omits sa 1
H // 
sa 9
: omits H
sa 15. Notes
Flesh
8 [
]: Although there was presumably room for this reading in the preceding line (cf. the length of lines 2, 3 and 8), there is a vertical stroke just to the right of the tear of the parchment that I take to be the second hasta of nu.11 It is unfortunate that part of the actual ἑρμηνεία (on both flesh and hair) is lost, although we can make a couple observations. First, this comment is a one-liner, since subsequent text would be visible (cf. hair side). Second, the statement begins with ‘It is necessary for you [sg.] to …’ (
), which would be completed with an infinitive. Based on the context of the Johannine passage quoted, a reasonable reconstruction would be (‘It is necessary for you to believe in the name’). While this is admittedly only a guess, we might note that the verb πιστεύω/ occurs in the ἑρμηνεία in P.Berlin inv. 11914. It is also worth noting the use of the second singular masculine (), which is found in four other ἑρμηνεῖαι (P.Monts. Roca 83, P.Berlin inv. 11914, P.Vindob. G 36102, Paris, BnF Copte 156). Papini drew attention to this common feature (second singular masculine) in the Coptic of P.Berlin inv. 11914 and Florence, Antinoe Copte 22.Footnote 26
]: Although there was presumably room for this reading in the preceding line (cf. the length of lines 2, 3 and 8), there is a vertical stroke just to the right of the tear of the parchment that I take to be the second hasta of nu.
), which would be completed with an infinitive. Based on the context of the Johannine passage quoted, a reasonable reconstruction would be 
(‘It is necessary for you to believe in the name’). While this is admittedly only a guess, we might note that the verb πιστεύω/
occurs in the ἑρμηνεία in P.Berlin inv. 11914. It is also worth noting the use of the second singular masculine (
), which is found in four other ἑρμηνεῖαι (P.Monts. Roca 83, P.Berlin inv. 11914, P.Vindob. G 36102, Paris, BnF Copte 156). Papini drew attention to this common feature (second singular masculine) in the Coptic of P.Berlin inv. 11914 and Florence, Antinoe Copte 22.Hair
5–6 The phrase
translates the Greek phrase καὶ οὐκ ἔρχεται πρὸς τὸ φῶς. Its complete omission in sa 1 can probably be attributed to parablepsis facilitated by homoeoteleuton, since occurs twice in close proximity (see transcript above).6
]: I have not followed Horner here in reconstructing , as the length of this line suggests against this reading. Coptic uses both and interchangeably in purpose or result clauses; here it is translating ἵνα.7
: This phrase corresponds to the Greek phrase ὅτι πονηρά ἐστιν, which is a variant reading found in the text of John 3.20 in several Greek manuscripts, notably P.Bodmer ii (P66). According to Bruce Metzger, the reading is a ‘natural expansion’ derived from the previous verse.Footnote 27 Thus, our manuscript, following the wider Coptic textual tradition, includes the phrase.14–15 The ἑρμηνεία consists of two lines and is difficult to reconstruct. Based on what is preserved, perhaps we have an imperative with
followed by the result of the action – for example, … (‘Obey him and you will …’). This is, however, an example and nothing more. Too little text remains for any plausible reconstruction to be made. It should be noted, however, that the conjunctive () is a second person singular (cf. recto). The last two words may be something like (cf. John 11.37), but there are certainly other possibilities.
translates the Greek phrase καὶ οὐκ ἔρχεται πρὸς τὸ φῶς. Its complete omission in sa 1 can probably be attributed to parablepsis facilitated by homoeoteleuton, since 
occurs twice in close proximity (see transcript above).
]: I have not followed Horner here in reconstructing 
, as the length of this line suggests against this reading. Coptic uses both 
interchangeably in purpose or result clauses; here it is translating ἵνα.
: This phrase corresponds to the Greek phrase ὅτι πονηρά ἐστιν, which is a variant reading found in the text of John 3.20 in several Greek manuscripts, notably P.Bodmer 
followed by the result of the action – for example, 
… (‘Obey him and you will …’). This is, however, an example and nothing more. Too little text remains for any plausible reconstruction to be made. It should be noted, however, that the conjunctive (
) is a second person singular (cf. recto). The last two words may be something like 
(cf. John 11.37), but there are certainly other possibilities.6. Concluding Remarks
P.CtYBR inv. 4641 is important for a number of reasons. First, it extends our knowledge of Sahidic manuscripts of John's Gospel. As mentioned above, it has already been assigned the call number ‘sa 972’ in K. Schüssler's Biblia Coptica, and just prior to submitting the final draft of this article, Siegfried Richter of Münster informed me that it has been registered in the official list of Coptic New Testament manuscripts (Schmitz–Mink–Richter) with the SMR number ‘sa 402’.Footnote 28 It will therefore come to play a role in New Testament textual criticism.Footnote 29
Second, it enriches our knowledge of ἑρμηνεία manuscripts of John, becoming the first known example of a Johannine ἑρμηνεία manuscript written solely in Coptic. As such, P.CtYBR inv. 4641 provides firm evidence that the production of these enigmatic manuscripts of John took place within Coptic Christian communities. To date, the biblical lemmata of Greco-Coptic ἑρμηνεία manuscripts of John exist only in Greek, which may indicate that Coptic Christians adopted the practice from their Greek-speaking predecessors with whom the practice originated. But this prompts the question: how extensive were these manuscripts within Coptic Christianity in Late Antiquity? If we base our reasoning on the evidence of Coptic-only ἑρμηνεῖαι, then the answer would inevitably be that this textual phenomenon was not very popular in circles where Coptic was the primary language.Footnote 30 On the other hand, the fact that we have multiple Greco-Coptic ἑρμηνεία statements demonstrates that the practice did take place in communities in which both Coptic and Greek were presumably used simultaneously.Footnote 31 Thus, the bilingual character of other ἑρμηνεία manuscripts is significant, as Cirafesi has shown.Footnote 32
Many questions remain, however, not only for our manuscript but ἑρμηνεία manuscripts of John in general. For one, it is still not clear whether these texts were created for private or public reading. If they served the liturgical and catechetical needs of individual communities (as Cirafesi argues), then how did the anagnostes proceed with both the reading of John and the ἑρμηνεῖαι? And why was John the text of choice? It is true that ἑρμηνεῖαι occur later in other biblical books (e.g. in Mark in Codex Bezae), but the evidence suggests the practice was first applied to the text of John's Gospel.Footnote 33 It seems that ἑρμηνεῖαι were used early on for the purpose of divination just as they were in later manuscripts like Codex Bezae, and Wilkinson has provided the best explanation of the oracular system to date.Footnote 34 We may never have all the answers with respect to these and other questions, but it is remarkable that space was given to these oracular comments alongside scriptural citations in a composite form. Presuming that the various ἑρμηνεῖαι were produced in scribendo and not from an exemplar (although this possibility cannot be ruled out), their presence alongside scripture demonstrates that scribes were actively engaged in the process of bibliomantic interpretation. That is, the oracular statements were not afterthoughts but part of the process of manuscript production. In any case, future studies on the ἑρμηνεία manuscripts of John will have to take questions such as the ones raised here into consideration, and P.CtYBR inv. 4641 will certainly be part of those studies.
Figure 1. P.CtYBR inv. 4641 – Flesh
Figure reproduced with kind permission of Beinecke Rare Book and Manuscript Library and supplied by Yale University.
Figure 2. P.CtYBR inv. 4641 – Hair
Figure reproduced with kind permission of Beinecke Rare Book and Manuscript Library and supplied by Yale University.
. See also 
